The endosomal protein-sorting receptor sortilin has a role in trafficking α-1 antitrypsin

Abstract

Up to 1 in 3000 individuals in the United States have α-1 antitrypsin deficiency, and the most common cause of this disease is homozygosity for the antitrypsin-Z variant (ATZ). ATZ is inefficiently secreted, resulting in protein deficiency in the lungs and toxic polymer accumulation in the liver. However, only a subset of patients suffer from liver disease, suggesting that genetic factors predispose individuals to liver disease. To identify candidate factors, we developed a yeast ATZ expression system that recapitulates key features of the disease-causing protein. We then adapted this system to screen the yeast deletion mutant collection to identify conserved genes that affect ATZ secretion and thus may modify the risk for developing liver disease. The results of the screen and associated assays indicate that ATZ is degraded in the vacuole after being routed from the Golgi. In fact, one of the strongest hits from our screen was Vps10, which can serve as a receptor for the delivery of aberrant proteins to the vacuole. Because genome-wide association studies implicate the human Vps10 homolog, sortilin, in cardiovascular disease, and because hepatic cell lines that stably express wild-type or mutant sortilin were recently established, we examined whether ATZ levels and secretion are affected by sortilin. As hypothesized, sortilin function impacts the levels of secreted ATZ in mammalian cells. This study represents the first genome-wide screen for factors that modulate ATZ secretion and has led to the identification of a gene that may modify disease severity or presentation in individuals with ATZ-associated liver disease.

Figure 1

Human AT expressed in yeast resides primarily in the ER. (A) ATM and ATZ expression in wild-type yeast was induced by transfer of transformed cells to growth medium lacking methionine. Cell extracts were taken at the indicated times and AT levels were assayed by immunoblot. G6PDH was used as a loading control. The two top bands represent the ∼55- and 53-kDa species. (B) Indirect immunofluorescence microscopy of BiP and AT indicates that ATM and ATZ are primarily localized to the ER (visible as a perinuclear ring). The relative location of the nucleus is shown by DAPI staining. (C) Live-cell fluorescence microscopy of AT-GFP and Sec63-mCherry supports the ER residence of ATM and ATZ. (D) To determine whether ATM and ATZ reside within a membrane compartment, cell lysates were separated into an ER-membrane-enriched pellet fraction (P) and a nonmembrane supernatant (S), and levels of AT, the ER lumenal protein BiP, and the cytosolic protein G6PDH were assayed by immunoblot. (E) ER-enriched microsomes were incubated in the presence or the absence of trypsin and TritonX-100 as indicated, and the levels of AT, BiP, and the ER membrane protein Sec61 were assayed by immunoblot. Arrows indicate the position of the doublet that indicates an intermediate BiP degradation product formed after trypsin treatment. (F) A pulse-chase analysis was used to establish N-linked glycosylation of AT. ³⁵S-labeled-AT was immunoprecipitated and treated with either Endo H or incubated with buffer before being analyzed by SDS-PAGE. “V” indicates vector control samples (which lack AT). In Figures 1 and 2 an asterisk indicates an AT species that was not found within the ER lumen.

Figure 2

ATZ is less efficiently secreted and is less detergent-soluble than ATM. (A) Wild-type yeast expressing ATM (M) or ATZ (Z) were grown to log phase and cell lysates (intracellular; IC) and conditioned media (extracellular; EC) were prepared and treated in the presence (+) or the absence (−) of Endo H before being analyzed by immunoblot. The large brace indicates high-molecular-weight species corresponding to hyperglycosylated AT. Arrows indicate the positions of the 95- and 55-kDa molecular mass standards for comparison of the EC and IC species. (B) ATM- or ATZ-expressing cells were converted to spheroplasts and were lysed with Triton X-100. The extracts were then separated into total (T), soluble (S), and insoluble (P) fractions by centrifugation, and AT and the soluble protein G6PDH were analyzed by immunoblot.

Figure 3

ATZ levels and secretion can be assayed by colony immunoblot. (A) Wild-type and vps30Δ cells expressing ATM (M) or ATZ (Z) were grown to saturation in the presence of methionine and diluted to an OD₆₀₀ of 1.0, and colony immunoblots were performed. After immunoblotting for AT, the same blot was stripped and immunoblotted for the cytosolic protein G6PDH. (B) Quantitation of AT immunoblots from A. Three transformants were quantitated for each sample, and the mean background-corrected signal intensity is represented. Error bars indicate the standard deviation of the mean. (C) Mutant strains from plate 109 from the yeast deletion mutant collection were transformed with pATZ, and colony immunoblots were performed. Control spots are indicated by numbers: 1: wild type/ATM; 2: wild type/ATZ; 3: vps30Δ/ATZ. Solid boxes indicate the position of the vps30Δ strain within the plate, and the dashed boxes indicate the position of met8Δ strain, which is mutated for an enzyme involved in siroheme synthesis, and hence is defective for sulfite reductase activity and methionine biosynthesis. A small amount of met8Δ growth was supported by the residual methionine carried over from the starter culture during replica spotting. In A–C, IC + EC (total) indicates lysed colony immunoblots and EC (secreted) indicates unlysed colony blots. In C, the EC membrane was imaged using a more sensitive chemiluminescence reagent than the IC + EC membrane to allow comparison.

Figure 4

AT is trafficked to and degraded in the yeast vacuole. (A) Immunofluorescence microscopy of AT and BiP and nuclear localization by DAPI staining is shown in pep4Δ cells expressing ATM or ATZ. The position of the vacuole is visible in the DIC images and corresponds to the position of the majority of the AT signal. (B) AT pulse-chase analysis in wild-type and pep4Δ cells expressing ATM or ATZ. The large brace indicates the high-molecular-weight AT species that accumulate in the pepΔ strain. Arrows indicate an unglycosylated AT precursor (fastest migrating species) and two glycosylated forms of AT. (C) Quantification of the pulse-chase data from B. Three transformants were examined, and the averaged amount of AT remaining at each time point was expressed relative to the amount of AT present at the zero time point. Error bars represent the standard deviation. Asterisks indicate the time points at which there was a statistically significant (P < 0.05) difference in relative AT levels between wild-type and pep4Δ cells. (D) AT was immunoprecipitated from the samples examined in B and was incubated in the presence (+) or the absence (−) of Endo H and then analyzed by SDS-PAGE. The large brace indicates the high-molecular-weight AT species that accumulate in the pep4Δ strain. Arrows indicate an unglycosylated AT precursor (fastest migrating species) and two glycosylated forms of AT.

Figure 5

Overexpression of the human homolog of Vps10 sortilin in rat hepatoma cells affects intracellular and extracellular AT levels. (A) ATM-FLAG and ATZ-FLAG levels are decreased in McArdle RH-7777 cells that overexpress human wild-type sortilin (S) in comparison to cells that express the Sort.LAYA mis-sorting mutant version of sortilin (L). Cells stably expressing sortilin or Sort.LAYA were transiently transfected with pATM-FLAG, pATZ-FLAG, or the empty vector pRc/RSV (“V”). After 24 hr, cells were labeled for 20 min with ³⁵S-amino acids, followed by a 20-min chase. AT levels in cell lysates (IC) and media (EC) were analyzed by immunoprecipitation with anti-FLAG antibody and subjected to SDS-PAGE analysis. The low levels of ATZ-FLAG that were secreted after 20 min are detectable by enhancement of radiograph contrast and brightness (bottom). (B) Quantification from five independent replicates of this experiment is presented as a scatter plot showing individual sortilin vs. Sort.LAYA comparisons as paired data points. Each individual data point is represented by a dot, and comparisons across adjacent positions in a gel are connected by a line. The averaged data and statistical analysis for this experiment are shown in Table 3.